[01:45] <gerbilcabbagehai> I have my Desktop box server running. 12.04. How do I log into it via sftp from my client laptop? I'm already in via ssh, but I don't know how to use ftp
[01:45] <sarnold> ftp != sftp
[01:45] <gerbilcabbagehai> i know
[01:46] <sarnold> which one are you interested in using? :)
[01:46] <gerbilcabbagehai> sftp
[01:46] <gerbilcabbagehai> on my client box i am using filezilla
[01:46] <sarnold> does filezilla support sftp? I've never seen it before..
[01:46] <gerbilcabbagehai> apparently
[01:47] <sarnold> if it takes URLs, try something like sftp://hostname/pathname
[01:47] <gerbilcabbagehai> http://filezilla-project.org/client_features.php
[01:47] <gerbilcabbagehai> ok thanks ill try that
[01:49] <sarnold> ah, those screenshots don't look very amenable to URL syntax
[01:49] <gerbilcabbagehai> it worked
[01:49] <sarnold> oh! pleasant surprise. :)
[01:49] <gerbilcabbagehai> filled in host, selected sftp, supplied username and password for ssh
[01:49] <gerbilcabbagehai> thanks for your help :)
[01:50] <sarnold> you're welcome :)
[01:54] <gerbilcabbagehai> if I want to write/overwrite when I am logged in via sftp (filezilla), the permissions are denied. What do I change on the server to continue modifying my files from afar, but keeping them safe from not-logged in users?
[01:55] <sarnold> gerbilcabbagehai: what are the permissions on the file and the file's containing directory?
[01:56] <gerbilcabbagehai> you want cli format or filezilla gui format?
[01:56] <sarnold> I can read ls -l output far better :) but that's just me...
[01:56] <gerbilcabbagehai> im new to cli so i want to learn how to do that. do I cd to the directory?
[01:56] <sarnold> yes
[01:57] <sarnold> it isn't strictly necessary -- but it is easier to get the listing, and to read it, if you do.
[01:57] <gerbilcabbagehai> -rw-r--r--  1 root root  177 Sep 20 18:38 index.html
[01:57] <gerbilcabbagehai> drwxr-xr-x  3 root root 4096 Sep 20 23:59 movies
[01:57] <gerbilcabbagehai> drwxr-xr-x 58 root root 4096 Sep 21 08:48 usb
[01:58] <sarnold> do you want your user account to be able to re-write this file at will?
[01:58] <gerbilcabbagehai> yes for my account, no for others.
[01:59] <gerbilcabbagehai> there's only one account with a strong password, however others will have access TO it, but i dont want them to change anything inadvertantly
[01:59] <gerbilcabbagehai> TO it = the data.
[02:00] <sarnold> gerbilcabbagehai: you will need to grant directory write access to your user account as well as file write access to that file for your user account
[02:01] <sarnold> (both are required since different programs may use different mechanisms to modify the file -- some will edit that existing file, and others will put a _new_ file in place with that filename. It's a touch complicated that way...)
[02:01] <sarnold> gerbilcabbagehai: is this user in a group all to itself?
[02:02] <gerbilcabbagehai> Lets say my name is bob. bob is my login and username. bob is the only username on the server. I'll be the only one who changes things or overwrites things.
[02:02] <gerbilcabbagehai> im not quite sure what i want to ask
[02:02] <gerbilcabbagehai> for now, i only want bob to be able to modify anything
[02:02] <sarnold> then that's pretty easy: chown bob:bob . index.html ; chmod 755 . ; chmod 644 index.html
[02:03] <gerbilcabbagehai> what does chmod 755 mean?
[02:03] <sarnold> that changes the directory's permissions to be owned by user bob, group bob, then sets the directory permissions to allow bob complete control, everyone else gets read and traverse permissions (necessary for a web server running as a different user, probably www-data)
[02:04] <sarnold> then it changes the file index.html to be owned by user bob, group bob, and allows bob to read and write the file, and everyone else can read the file (also necessary for the web server to be able to serve it up to clients)
[02:04] <gerbilcabbagehai> so what does having root as owner do to the file index.html?
[02:04] <gerbilcabbagehai> why should bob own it?
[02:05] <gerbilcabbagehai> e.g. sarah / guest can see and read bob's index, but only bob can modify?
[02:06] <sarnold> if root owns the file, then bob has use 'sudo' to modify the file, which gets annoying
[02:06] <gerbilcabbagehai> true :)
[02:06] <sarnold> if bob owns the file, then bob can update the file without having to use 'sudo'
[02:07] <gerbilcabbagehai> hence if i own that file, i login via filezilla and permissions will be allowed because I own that file, and filezilla cannot pass sudo (i think)
[02:07] <sarnold> sarah will get the "other" permissions -- the last digit in the permissions: 755 and 644
[02:07] <sarnold> right, filezilla won't be able to do sudo, especially because the sftp protocol doesn't make a provision for that :)
[02:07] <gerbilcabbagehai> gotcha. thanks for the explaination
[02:08] <gerbilcabbagehai> now to try it lol
[02:08] <sarnold> hehe
[02:08] <gerbilcabbagehai> new meaning to the term, "man up"
[02:09] <gerbilcabbagehai> you know...many things i think are deliberate in linux
[02:09] <sarnold> one hopes :)
[02:09] <gerbilcabbagehai> is there a $kill bill?
[02:10] <sarnold> back in the day, the 'make' command would print "make love, not war" if you typed "make love"...
[02:10] <gerbilcabbagehai> AHAHAHA
[02:11] <gerbilcabbagehai> how do i enable logging on the server?
[02:11] <sarnold> gerbilcabbagehai: an old old joke stuffed in the tunefs manpage back in the eighties: http://unixhistory.livejournal.com/1808.html
[02:12] <sarnold> gerbilcabbagehai: what operations do you want logged?
[02:12] <gerbilcabbagehai> everything?
[02:12] <sarnold> you don't want that :) that's way too much data...
[02:12] <gerbilcabbagehai> um...let me rephrase that
[02:13] <gerbilcabbagehai> any user who logs in and does something
[02:13] <gerbilcabbagehai> from where and as much detail as possible about them
[02:14] <gerbilcabbagehai> all interactions that my username and any other initiates
[02:14] <sarnold> well, that's still pretty open...
[02:14] <sarnold> gerbilcabbagehai: first, try "sudo less /var/log/auth.log" -- see if that's sufficient
[02:14] <gerbilcabbagehai> am i going to regret this? lol
[02:14] <sarnold> no, 'less' is a simple file viewer
[02:14] <sarnold> just hit 'q' to quit
[02:15] <gerbilcabbagehai> i mean am i going to regret asking for so much
[02:15] <sarnold> you may also like to try the "last" command
[02:18] <gerbilcabbagehai> chown bob:bob . index.html ; chmod 755 . ; chmod 644 index.html ... results in
[02:18] <gerbilcabbagehai> chown: changing ownership of `.': Operation not permitted
[02:18] <gerbilcabbagehai> chown: changing ownership of `index.html': Operation not permitted
[02:18] <gerbilcabbagehai> what am i typing incorrectly?
[02:18] <sarnold> oh bother. my fault.
[02:19] <gerbilcabbagehai> ./index...?
[02:19] <sarnold> gerbilcabbagehai: you'll need to run those commands via 'sudo'. If you run 'sudo -s' it'll start a root shell for you...
[02:19] <sarnold> gerbilcabbagehai: I'm being summoned for dinner :) so I'm off for the night. have fun!
[02:20] <gerbilcabbagehai> thanks :)
[02:27] <Bluemalkin> hi - does anyone know hot to do an approx apt cache refresh in 12.04 since the approx-update has been removed ?
[02:29] <gerbilcabbagehai> does root have more permissions than username? i.e. if I mistakenly make /var/www permissioned to USERNAME only, can root still access?
[03:19] <linocisco> hi all, HOw can I save jpg/pdf or scanned documents in database.?
[03:20] <linocisco> not as only link. Just physical image file
[03:54] <linocisco> hi all
[04:02] <linocisco> is there any LDAP like settopbox so that windows clients can authenticate and see their designated folder and their roaming home folders?
[07:41] <cluelessperson> Does anyone know how to restrict an SFTP user into only their home directory?
[07:44] <RoyK> cluelessperson: use rssh to disallow normal login and read up on chrooting
[07:45] <cluelessperson> RoyK: Is there an easier way to chroot?
[07:45] <cluelessperson> RoyK:  I've got rssh working
[07:50] <RoyK> cluelessperson: dunno, try google
[07:56] <cluelessperson> RoyK: I've spent way too much time on google without results.
[08:11] <Insomniac11> Hello, I need help troubleshooting an Ubuntu 12.04 Server-amd64 installation. I keep getting the errors listed in this pastebin: http://pastebin.com/frpZkMAt I also looked in the last 20 lines of the syslog and can't make sense of it, I attatched that to the pastebin as well for all to see. The pastebin has more detailed information as well.
[08:18] <ikonia> Insomniac11: it looks like you have errors that have already scrolled off the screen
[08:19] <Insomniac11> Hmm, I thought they were just wordwrapped down tot he next line??
[08:23] <ikonia> Insomniac11: your first line is saying "you may want to correct this problem" - so there is obviously errors above that
[08:24] <Insomniac11> Well, I read the last 20 lines were pernient somewhere where someone had a similar problem, and I didn't know how to export the entire syslog into a separate file from within the terminal to an ourside source.
[08:26] <Insomniac11> I'm sorry, I'm not very familiar with all of this. I will try to get you more information.
[08:29] <Insomniac11> I can give you access to my VNC? Then you can see the whole log
[08:29] <Insomniac11> as it is on my screen
[08:29] <ikonia> no thank you
[08:30] <Insomniac11> Any suggestions for forward progress?
[08:32]  * Insomniac11 shrugs
[08:32] <Insomniac11> Maybe someone can help in the #ubuntu channel. Thanks anyways ikonia.
[09:05] <Dulcin> I just accidentally chown and chgrp everything in /var/ to my user, I've put it back to root - but is that how is used to be?
[09:25] <lordievader> Good morning.
[09:28] <jamespage> Dulcin, almost certainly not
[09:29] <RoyK> cluelessperson: first hit on google http://www.cyberciti.biz/tips/howto-linux-unix-rssh-chroot-jail-setup.html
[09:30] <RoyK> cluelessperson: good choice of nick, btw ;)
[09:31] <cluelessperson> RoyK: So I've followed the directions, and it doesn't work
[09:31] <RoyK> cluelessperson: then you're doing something wrong
[09:32] <cluelessperson> RoyK: http://unixhelp.ed.ac.uk/CGI/man-cgi?sshd_config+5
[09:33] <cluelessperson> according to that I can chroot and restrict sftp users without other dependencies, that I must set up chroot to be root owned but not rw by any other group.
[09:33] <cluelessperson> which means chmodding the chroot directory
[09:33] <cluelessperson> problem being, it's automounted by fstab
[09:35] <cluelessperson> so, I need to dictate in fstab the umask= something
[09:36] <cluelessperson> RoyK: do you know off the top of your head to only allow root access to either r or w ?
[09:36] <cluelessperson> mode
[09:37] <RoyK> cluelessperson: man setfacl
[09:37] <RoyK> !acl
[09:49] <cluelessperson> RoyK: Oi.  I'm getting closer.  Now I've only got a permission denied
[09:54] <cluelessperson> RoyK: Okay, so now the issue is what it's not restricting the user to their home directory
[09:55] <RoyK> cluelessperson: I've answered that question
[09:55] <cluelessperson> do chroot?  I did though
[09:56]  * RoyK gives up and finds something better to do than talking to rather clueless people
[09:56] <cluelessperson> RoyK: This isn't obvious.
[09:57] <cluelessperson> RoyK: First fstab has to mount it with appropriate permissions, which I have no idea what those are, supposedly only root read and write access.
[09:57] <cluelessperson> second, this user is configured to use the rssh shell, denied scp access.
[09:57] <cluelessperson> their home in /media/sftp/guest/
[12:13] <SpamapS> jamespage: did you see that ceph 0.48.2 was released on Wednesday?
[12:41] <roaksoax> Daviey: https://bugs.launchpad.net/ubuntu/+bug/1054022
[13:02] <doko> Daviey, zul: is the maas-provision demotion ok for quantal?
[13:02] <zul> doko: no idea
[13:15] <Daviey> doko: I wouldn't worry about it.  It's either going to be plainly removed, or turned into a transitional package shortly.
[13:15] <doko> Daviey, ok, demoting. transitional packages can say in universe
[13:57] <_val_> Hi everyone. Using LTS 12.04. After a reboot I got to (initramfs). I fixed this issue once without using a live CD or else. Now how would I set the rootpat?  I did  set root=hd0,0   returned nothign. How do I beet the preivous kernel?
[14:00] <stgraber> hallyn: thoughts on https://github.com/lxc/lxc/issues/4 ?
[14:03] <doko> Daviey, the netcf binary pops up on the demotion list, although the source was promoted this cycle. is the demotion correct?
[14:10] <jamespage> _val_: hit escape on boot - you should get a menu of choices...
[14:18] <Daviey> doko: probably not
[14:19] <doko> please seed it
[14:22] <hallyn> stgraber: nothing we can do about that, it's kernel enforced
[14:23] <hallyn> stgraber: that's where optionalling doing chroot instaed of pivot_root would come in
[14:23] <_val_> jamespage: does not help.
[14:23] <hallyn> stgraber: but that won't fix certain things like /proc/maps and /proc/mounts
[14:23] <jamespage> _val_, does the same thing? hmmm
[14:24] <jamespage> the server iso has a rescue mode
[14:24] <_val_> I've set up a tunnel for the ipmi interface otherwise I can't reach the server. Either the Esc key is not working.. and when typing 'boot' it just says. Command not found
[14:24] <doko> SpamapS, jamespage: you seem to care about ceph. python-ceph is schedule for demotion. is this intended?
[14:24] <jamespage> doko, no
[14:24] <hallyn> stgraber: so we *could* recommend a particular ugly combination of virtual devices set up for /var/lib/lxc/ which would make MS_SHARED work, i *think* we can make it work...  fugly though
[14:24] <jamespage> let me seed it
[14:24] <doko> thanks
[14:25] <hallyn> stgraber: but the shared setting has to be right for both the device and it's parent iirc
[14:25] <hallyn> stgraber: (requirements are documented in fs/namespace.c)
[14:26] <hallyn> stgraber: (biab)
[14:26] <doko> Daviey, cloud-initramfs-rescuevol demotion?
[14:28] <stgraber> hallyn: so I guess there's no way of setting the initial bind mount as PRIVATE so it doesn't bring the whole tree with it?
[14:29] <hallyn> stgraber: not sure what you mean;  but a single mount won't do it
[14:29] <hallyn> we'd probably want to make a tmpfs mount and then mount /var/lib/lxc onto there
[14:29] <hallyn> stgraber: it's worth playing with.
[14:30] <stgraber> hmm, yeah... ugly, so not something we should be doing by default. Maybe get the magic in a pre-mount hook that they can use on systems with systemd
[14:31] <hallyn> that's an idea
[14:31] <jamespage> doko, done
[14:31] <Daviey> doko: Unless you are in a hurry for this, i'll process cloud related demotions on Monday.
[14:31] <hallyn> or pre-start
[14:31] <hallyn> i was thinking just recommending a best practice for setting up lxc at install time
[14:31] <doko> Daviey, that's fine. just doing my +1 maint
[14:32] <hallyn> stgraber: it would be fun to play with, but give we have our own priorities and time constraints, it seems like something we should guide one of our fine fedora friends through
[14:35] <stgraber> hallyn: yeah, not worth spending time working around weird systemd "features"
[14:36] <hallyn> stgraber: looking back through, current->fs->root->mnt-parent must not be MNT_SHARED - that's probably the hardest one to handle
[14:37] <jamespage> SpamapS, no - I'd missed that
[14:38] <jamespage> ceph 0.48.2 that is....
[14:38] <doko> Daviey, the last one is python3-requests
[14:38] <hallyn> stgraber: libvirt-lxc will of course have the same issue .  so really i expect it to be fixed in fedora pretty quickly
[14:43] <jamespage> SpamapS, release notes and commit history in github look reasonable
[14:43] <jamespage> http://ceph.com/releases/v0-48-2-argonaut-stable-update-released/
[14:46] <jamespage> I'll give it a test over the weekend and upload early next week
[14:50] <SpamapS> jamespage: I already started actually
[14:51] <SpamapS> jamespage: I was wondering how you've been creating the packages. I cloned git, checked out the tag, and then archive/uupdate'd from that
[14:51] <jamespage> SpamapS, use the watch file in the package - it does the right thing
[14:51] <jamespage> uscan --rename --verbose
[14:51] <jamespage> I think
[14:51] <SpamapS> jamespage: it produced a really gigantic diff
[14:52] <SpamapS> jamespage: all the generated files added and stuff
[14:53] <jamespage> SpamapS, really - odd - thats how I did it last file
[14:54] <jamespage> you are using the official ubuntu packaging branch as source right?
[14:55] <jamespage> grrr - now I remember
[14:55] <jamespage> bzr is crap at merging a tarball with a debian subfolder
[14:55] <jamespage> it should discard it IMHO
[14:55] <SpamapS> jamespage: I gave up on the bzr way rapidly
[14:56] <SpamapS> jamespage: I didn't spend much time investigating.. the git tag produced an expected diff
[14:58] <hallyn> plars: wow, this is messed up
[14:59] <hallyn> plars: i installed a precise server.  installed openssh-server.
[14:59] <hallyn> upgraded.  rebooted.  i have the console that doesn't update.
[14:59] <hallyn> plars: but i log in over ssh and watch ps output,
[14:59] <plars> hallyn: yeah, that's exactly what I'm seeing
[14:59] <hallyn> while i type my username /pwd onto the dead console.  then say 'vi /tmp/ab'.  'login --', bash, and vi /tmp/ab show up in ps -ef output!
[14:59] <jamespage> SpamapS, I don't have time todo it now - if you do update it please can you update the arch for python-ceph - I think it should be 'all'
[15:00] <jamespage> otherwise I'll pickup +2 days
[15:00] <hallyn> plars: so the consoles are working and getting input.
[15:00] <plars> hallyn: I haven't messed with the networking enough to sort out how to get a connection to it, what's incantation for that?
[15:00] <plars> hallyn: so it's not truly frozen, that's good at least
[15:00] <hallyn> plars: yeah this looks like a bug in the vnc output in qemu-kvm
[15:01] <hallyn> plars: not sure which incantation you mean.  i had the vm bridged to my br0, and watched dhcp server logs to get the ip address, and ssh'd to it
[15:02] <hallyn> and, if i hit 'send ctrl-alt-f2' in the vnc client menu, and log back in, i get a second session :)
[15:02] <hallyn> all right, i can file a bug on this.  thanks plars !
[15:02] <plars> hallyn: I'd be happy to file it, I just have no idea where to put it
[15:02] <SpamapS> jamespage: got it. Will do
[15:02] <hallyn> plars: oh i'm just going to qemu-devel.  feel free to email yoruself if you like.
[15:02] <hallyn> plars: now i'm wondernig if non-upgrade reboots will do it too
[15:03] <hallyn> plars: change of locale, biab
[15:07] <jamespage> SpamapS, ta
[15:12] <plars> hallyn: jibel just made an awesome discovery, booting with -vga std seems to make it happy
[15:15] <hallyn> plars: oh the default is cirrus?  bah.
[15:16] <hallyn> plars: i suppose it could even be the kernel driver in quantal
[15:16] <plars> hallyn: that's what it sounds like then...
[15:16] <hallyn> plars: so before oing to qemu-devel i'll file a bug in lp :)  to track our discoveries
[15:17] <hallyn> plars: would you mind doing a bunch of reboots in both precise and quantal vms and seeing if either/both get the no console after a regular reboot?
[15:17] <hallyn> plars: i'm also curious whether rebooting (without hard-power-off) can also fix the console
[15:17] <plars> hallyn: I'm doing it on the precise one right now, I can check in a bit on quantal
[15:18] <hallyn> plars: awesome, thanks.  i'll start the lp bug
[15:20] <hallyn> plars: say, are you using vnc, or are you using SDL?
[15:20] <hallyn> if you're using SDL then that makes it much less likely to be a qemu bug (as that's a different output system having the same bug as vnc)
[15:21] <plars> hallyn: sdl
[15:22] <hallyn> plars: interesting.  thanks
[15:41] <Queops> Hello! https://help.ubuntu.com/12.04/serverguide/postfix.html => Configuring SASL, I can't find what it wants me to edit on /etc/dovecot/dovecot.conf
[16:04] <fullstop> Hey.. I have a 12.04 server, and it appears to be trying to mount -a before starting lvm.
[16:05] <fullstop> any idea how to get it to delay or change the order?
[16:05] <fullstop> switch to uuid rather than path?
[16:09] <xnox> fullstop: what's the actual problem why you want it not do that?
[16:09] <xnox> at boot we have udev running and adding lvms.
[16:10] <xnox> fullstop: you probably want to modify your /etc/lvm.conf if you don't want thousands of them =)
[16:10] <fullstop> at boot, it states that the device is not ready and asks if I wish to skip it.
[16:10] <hallyn> ahs3: self-tests all passed with http://people.canonical.com/~serge/netcf022/ . not sure if i should test an ubuntu version (with libvirt) before asking for the debian one to be pushed...
[16:10] <xnox> fullstop: and what is the device in question and why is it not ready?
[16:11] <ahs3> hallyn: couldn't hurt, y'know...
[16:11] <fullstop> The device is just a logical volume.  I assumed that it was not ready because lvm2 had not been started yet.
[16:11] <xnox> fullstop: please file a bug against mountall with your stab, fdisk and lvs & vgs.
[16:12] <xnox> fullstop: well, actually lvm2 in that case
[16:12] <hallyn> ahs3: :)
[16:12] <xnox> fullstop: udev rules should find and activate all VGs rather quickly
[16:12] <fullstop> I find it hard to believe that I'm the first to run into this
[16:12] <fullstop> http://pastebin.com/4Sv8cKvv
[16:13] <xnox> you have raid as well, so it's not just lvm.
[16:13] <fullstop> No, there is no raid.
[16:14] <fullstop> In fact, this setup only has one drive right now.
[16:14] <xnox> why is it assembling "MD arrays" - mdadm -> RAID arrays.
[16:14] <xnox> ?
[16:14] <fullstop> I'm guessing that it does that on all servers
[16:14] <xnox> no, it does not.
[16:15] <fullstop> I may have installed dmraid to get around another boot-time bug.
[16:15] <fullstop> I promise you, though, this system has just one drive.
[16:16] <xnox> fullstop: please file a bug against mountall and please attach: /etc/fstab, lvs, vgs
[16:16] <xnox> fullstop: something odd is going on though.
[16:17] <fullstop> okay
[16:17] <xnox> fullstop: ubuntu-bug mountall
[16:17] <xnox> should work on the server and collect relevant info.
[16:17] <xnox> and print an URL where to go in your browser to file a bug.
[16:18] <fullstop> got it
[16:19] <fullstop> I wish I could remember what package I installed to work around another boot-time problem.
[16:20] <sarnold> fullstop: /var/log/dpkg.log has some details when packages were installed; if you can recall vaguely when you installed that package, you may be able to find the log entries about it..
[16:20] <fullstop> sarnold: I looked through there already.  Nothing jumped out at me.
[16:22] <sarnold> fullstop: drat :)
[16:29] <fullstop> xnox: reported.  Here is a similar report: http://forums.debian.net/viewtopic.php?f=5&t=51146&start=0
[16:29] <xnox> fullstop: bug number?
[16:30] <fullstop> 1054190
[16:30] <xnox> fullstop: debian is not very relevant, as we do not use init scripts for lvm.
[16:30] <fullstop> noted
[16:30] <xnox> fullstop: bug 1054190 <---- this way I will get a link
[16:30] <xnox> tadah =)
[16:31] <fullstop> go uvirtbot
[16:32] <xnox> fullstop: ... and pvs
[16:32] <xnox> *sorry* =)
[16:33] <fullstop> We'll have to wait on that one.  The server is being rebooted and I'm remote.  Somebody has to walk up to it and press some keys.
[16:33] <xnox> fullstop: or are they on the partitions of the sda?
[16:33] <fullstop> I think that they are eating lunch right now.  :D
[16:33] <fullstop> They are partitions on /dev/sda
[16:34] <fullstop> I have something else..
[16:34] <xnox> and you have to skip to boot? what happens if you wait? this is disturbing.
[16:34] <fullstop> I think that it will wait forever.
[16:36] <fullstop> okay, added the output of parted -l
[16:36] <fullstop> be back in a bit
[16:54] <zul> jamespage:  i dont see how radosgw is failing
[17:01] <zul> jdstrand: so quantum
[17:12] <zul> jdstrand: cisco and niciria doesnt use rootwraper apparently
[17:13] <jdstrand> zul: ok. can you comment in the bug?
[17:13] <zul> jdstrand: i will
[17:13] <jdstrand> zul: ping me after and I'll ACK
[17:14] <zul> done
[17:16] <fullstop> xnox: http://pastebin.com/jggwtJ04
[17:16] <fullstop> I removed mdadm and rebooted.
[17:18] <fullstop> I've not mucked about with udev or anything like that, either.
[17:18] <fullstop> and the lvm2 init script has this in it: X-Start-Before:    checkfs mountall
[17:18] <fullstop> so I am kind of perplexed.
[17:28] <xnox> fullstop: interesting lvm2 init script you say?
[17:28] <xnox> fullstop: dpkg -S /etc/init.d/lvm2
[17:28] <xnox> or fullstop: dpkg -L lvm2
[17:29] <xnox> (we did have an accident of dropping clvm and reintroducing lvm2 init script, when we actually want to keep clvm and drop lvm2)
[17:29] <fullstop> lvm2: /etc/init.d/lvm2
[17:29] <fullstop> and the other dpkg command lists all of the files for the package.  I can paste that if need be
[17:29] <xnox> fullstop: on precise?
[17:29] <fullstop> yes
[17:30] <fullstop> Ubuntu 12.04 LTS \n \l
[17:30] <xnox> fullstop: dpkg -i lvm2
[17:30] <xnox> and fullstop: dpkg -L lvm2 | grep init
[17:30] <fullstop> dpkg-deb: error: `lvm2' is not a debian format archive
[17:31] <fullstop> http://paste.ubuntu.com/1219059/
[17:32] <xnox> fullstop: http://packages.ubuntu.com/precise-updates/amd64/lvm2/filelist
[17:32] <xnox> does not have lvm2 init script.
[17:32] <fullstop> This install is quite recent.
[17:32] <fullstop> and we've not gone around changing things willy-nilly.
[17:32] <xnox> fullstop: was it precise install or upgrade?
[17:32] <fullstop> install
[17:32] <fullstop> ubuntu-server
[17:33]  * xnox wonders where /etc/init.d/lvm2 came from
[17:33] <xnox> fullstop: what's the full lvm2 package version number?
[17:33] <xnox> dpkg -i lvm2
[17:33] <xnox> and
[17:33] <xnox> apt-cache policy lvm2
[17:34] <fullstop> http://paste.ubuntu.com/1219061/
[17:34] <fullstop> the first command wants to operate on a .deb package
[17:34] <xnox> fullstop: you are mixing debian & ubuntu
[17:35] <xnox> fullstop: if you want .95, just add quantal and pin lvm2 from quantal. But please note this is _not_ supported.
[17:35] <xnox> fullstop: and hence the boot problems you have.
[17:35] <fullstop> ಠ_ಠ
[17:35] <fullstop> I will find the person who added those repos
[17:35] <fullstop> sorry for wasting your time
[17:36] <xnox> fullstop: it's ok =) good luck.
[17:39] <skrite> hey all
[17:39] <xnox> fullstop: sounds like you want puppet management (which goes about overwritting "customizations" like these with approved/managed configs)
[17:39] <xnox> fullstop: or you can use landscape which does compliance management as well.
[17:39] <fullstop> I'll have a talk with a few people.
[17:46] <xnox> fullstop: you do want to check what other crap you got from testing, cause precise froze good ~4 months before debian did.
[17:46] <fullstop> Yes, going through that now.  Thanks.
[18:13] <hallyn> stgraber: pushed a single trivial fix to lp:ubuntu/precise-proposed/lxc.  NOt sure if you had other things to stage
[18:13] <stgraber> hallyn: nope, nothing on my side
[18:14] <hallyn> stgraber: hm, i'm loath to start an SRU for this one-liner
[18:14]  * hallyn tries to find other SRU-ables
[18:14] <stgraber> hallyn: oh, didn't see it was precise-proposed...
[18:15] <stgraber> hallyn: we have quite a few that probably should go as sru to precise
[18:15] <stgraber> hallyn: I'll have a look at what went into quantal, figure out what makes sense to push as sru
[18:15] <hallyn> hm, there's a bunch at https://bugs.launchpad.net/ubuntu/precise/+source/lxc
[18:24] <chris|> hallyn, would you consider #1047837 sru able?
[18:26] <hallyn> chris|: switching from sysvinit to upstart - probably not.  especially since there doesn't appear to be a bug introduced by using sysvinit, right?
[18:26] <hallyn> chris|: SpamapS ^ might know better
[18:28] <SpamapS> chris|: It says wishlist. We only do SRU's for High Impact bugs.
[18:28] <chris|> right, there's no bug, but worth a try :)
[18:28] <SpamapS> chris|: and really, whats the point of migrating zabbix to upstart?
[18:28]  * SpamapS heads out to see the shuttle fly-by
[18:29] <chris|> SpamapS, have upstart start/stop services/task when specific containers start/stop
[18:30] <chris|> ah, no, zabbix
[18:32] <chris|> well, for once the respawn is kind of nice, I can delay zabbix until my networking and database setup is actually finished
[18:32] <hallyn> stgraber: uh, jinkeys.  I mistyped when fixing the debian testing container creation.  that needs to be isc-dhcp-client, not server
[18:33] <stgraber> oops, indeed :)
[18:33] <hallyn> stgraber: guess i'll commit the fix to that to bzr for now
[18:38] <arosales1> utlemming: could you add appropriate release note entry for the work items completed on the https://blueprints.launchpad.net/ubuntu/+spec/servercloud-q-cloud-images BP ?
[18:39] <utlemming> arosales1: yup, will do
[18:39] <arosales> utlemming: thanks
[18:40] <arosales> utlemming: probably want to do the same for https://blueprints.launchpad.net/ubuntu/+spec/servercloud-q-awstools as most of the majority of the packaging work items look to be completed.
[18:41] <utlemming> arsoales: ack
[18:41] <arosales> utlemming: thanks
[19:16] <hallyn> stgraber: precise-proposed lxc tests fine, i'll push
[19:29] <stgraber> hallyn: ok
[19:30] <hallyn> hm, augeas in quantal seems broken
[19:45] <skrite> in virtualbox, can i run a virtual machine in the background, or like, from a screen or tmux session?
[20:07] <rjune> I put tasksel tasksel/first multiselect standard,edubuntu-desktop in my preseed file. It installed the standard system, but not edubuntu-desktop. Is that statement only able to specify a single target to install?
[20:13] <hallyn> ahs3: all right, that netcf (without the libvirt Conflicts) tests fine in quantal by itself and thorugh libvirt
[20:14] <ahs3> hallyn: sweet.  what's the url for the dsc file again?
[20:15] <hallyn> ahs3: http://people.canonical.com/~serge/netcf022/netcf_0.2.2-1.dsc
[20:16] <ahs3> hallyn: thx.  i'll try to get that in over the weekend
[20:18] <hallyn> ahs3: thanks!  ttyl
[20:41] <ninjix> anyone here ever used the python-libvirt to upload a virStorageVol?
[21:07] <robgault_> q
[22:07] <blizzow> okay, so I downloaded the iso for ubuntu server.  I ran the usb-creator-kde utility, formatted a 160 GB external USB drive.  then "installed" the iso on the drive via usb-creator-kde.  When I boot off the drive, the initial process is fine, but then during the hardware detection the thing just bitches and moans that it can't find the CDROM to continue the installation.  Anyone know a way around this?  I've tried to ln -sf /dev/sdf /dev/cdrom and that d
[22:07] <sarnold> blizzow: cut off at "and that d"
[22:08] <blizzow> "did not work."
[22:08] <sarnold> so close ;)
[22:08] <sarnold> blizzow: do any other virtual terminals show any debugging information?
[22:09] <blizzow> The debug console just shows it spinning at trying to detect cdroms or some similar verbiage.